Search Safety PM Project Intern (TikTok Safety Product) - 2026 Start
Core
Building and optimizing content safety systems using large language models (LLMs) to enhance review efficiency, risk control, and user trust.
Role type
Project Intern (Search Safety PM)
Builds
Scalable content safety solutions and LLM training infrastructure
Domain
AI Safety / Content Moderation / Large Language Models
Deliverable
production ML models
Required skills
Content sensitivity, analytical thinking, English writing and communication, LLM awareness, learning agility, ownership, execution ability
Preferred skills
Humanities/social sciences background, content annotation experience, prompt engineering experience
Technologies
Large Language Models (LLMs), labeling systems, taxonomies, prompt engineering
Responsibilities
Construct and maintain high-quality LLM training datasets and knowledge bases; design and iterate labeling systems and taxonomies; generate and polish massive LLM training data; draft and optimize prompts for model training and evaluation; conduct quality inspection and consistency calibration on datasets and prompts
Seniority
Intern